Unverified50% confidenceConceptExact time
Hierarchical retrieval (two-stage retrieval/recall-rerank architecture) is the current mainstream paradigm for large-scale vector retrieval: the first stage uses lightweight methods to quickly filter a candidate set, and the second stage uses heavier models for fine-grained scoring.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe two-stage retrieval strategy first uses fast vector retrieval to recall Top-N candidates, then uses a cross-encoder to precisely score those N candidates and select the final Top-K.79% similarUnverifiedRAG检索阶段采用两阶段策略：第一阶段向量检索采用双编码器架构进行粗排，第二阶段Rerank模型采用交叉编码器架构进行精排77% similarUnverifiedTwo-tower retrieval is the mainstream recall architecture for handling large-scale candidate sets in industrial recommender systems, using two independent neural networks to encode users and items respectively into low-dimensional dense vectors, with match quality measured by vector inner product or cosine similarity.69% similarUnverifiedFunnel-style layered evaluation architecture: the first layer uses cheap and fast semantic verifiers to pre-screen large volumes of traces, and the second layer passes ambiguous or high-value samples to a more powerful judge model for detailed evaluation.69% similarUnverifiedA tiered evaluation strategy uses small models or rules for initial filtering, deploys expensive large-model judges only at critical stages, uses caching to avoid redundant evaluations, and uses refined evaluation sets to reduce token consumption.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1025156API
curl https://kongchang.com/api/v1/knowledge/claims/1025156MCP
get_claim(id=1025156)